Screen gets stuck while running chromium os on beagleboard.

[saggu]

I am trying to port chromium os on beagleboard.
So far now i am able to boot up kernel (2.6.35.8) and from the logs i
can see it has mounted the rootfs also from mmc partition 2&3. Apart
from that there are some erroraneous messages also in kernel boot log.
It first starts up by flashing Chromium OS logo screen then it goes to
blank.

The complete logs are at :-
http://tinypaste.com/1918d4e3

The bootargs i passes are setenv bootargs 'console=ttyS2,115200n8
vram=12M omapfb.mode=dvi:1024x768MR-16@60 omapfb.debug=y
omapdss.def_disp=dvi'

Anyone please suggest me the solution -:
1. Do i need to add some additional chromium os specific patches in
kernel source code?
If yes please suggest me from where i can get these patches?
2. From the logs there seems to be some problem with framing buffers
-:

" Console: switching to colour frame buffer device 128x48
regulator_init_complete: incomplete constraints, leaving VDAC on "

what changes should i make in kernel config to overcome this problem?

regards,
Simranjeet Singh

[Jonathan Kliegman]

I assume from your comment about the logs that you're able to login and run around in a shell to debug things?  Basically that the system boots but the UI doesn't load?

Looking at the logs from X should be a good place to start.  As well as making sure you've got the right set of video drivers going.

Its also possible you'll need to modify the xorg config files.  The easiest way to iterate would be to make your image with the --noenable_rootfs_verification flag.  Then when you get on the box you can remount root readwrite (mount -o rw,remount /).  Then modify the Xorg config files (/etc/X11) or add new ones as needed.  Finally restarting X is easiest done by running either 'start ui' or 'restart ui' depending on if its running or not.

Once you can see X come up, /var/log/ui/ui.LATEST will show the output from Chromium as it tries to start up and would be the next spot to look.

In terms of framebuffer issues on Beagleboard I haven't played with it at all so can't give any specific help, just some general thoughts on where things go and what's gone wrong on other platforms.

--
Chromium OS discuss mailing list: chromium-...@chromium.org
View archives, change email options, or unsubscribe:
http://groups.google.com/a/chromium.org/group/chromium-os-discuss?hl=en

[Sonny Rao]

On Tue, Dec 20, 2011 at 7:25 AM, Jonathan Kliegman <kli...@chromium.org> wrote:
> I assume from your comment about the logs that you're able to login and run
> around in a shell to debug things?  Basically that the system boots but the
> UI doesn't load?
>
> Looking at the logs from X should be a good place to start.  As well as
> making sure you've got the right set of video drivers going.

Yeah /var/log/Xorg.0.log would be good to be able to tell if X even
started or not

> Its also possible you'll need to modify the xorg config files.  The easiest
> way to iterate would be to make your image with the
> --noenable_rootfs_verification flag.  Then when you get on the box you can
> remount root readwrite (mount -o rw,remount /).  Then modify the Xorg config
> files (/etc/X11) or add new ones as needed.  Finally restarting X is easiest
> done by running either 'start ui' or 'restart ui' depending on if its
> running or not.
>
> Once you can see X come up, /var/log/ui/ui.LATEST will show the output from
> Chromium as it tries to start up and would be the next spot to look.
>
> In terms of framebuffer issues on Beagleboard I haven't played with it at
> all so can't give any specific help, just some general thoughts on where
> things go and what's gone wrong on other platforms.
>
>
> On Tue, Dec 20, 2011 at 4:42 AM, saggu <sagg...@gmail.com> wrote:
>>
>> I am trying to port chromium os on beagleboard.
>> So far now i am able to boot up kernel (2.6.35.8) and from the logs i
>> can see it has mounted the rootfs also from mmc partition 2&3. Apart
>> from that there are some erroraneous messages also in kernel boot log.
>> It first starts up by flashing Chromium OS logo screen then it goes to
>> blank.
>>
>> The complete logs are at :-
>> http://tinypaste.com/1918d4e3

From the unhandled prefetch abort messages, it looks like something is
crashing, but I can't tell what.
Can you enable fatal signal printing so that we can at least see which
process is dying?

To do that either edit sysctl.conf and add
"kernel.print-fatal-signals = 1"
or echo 1 > /proc/sys/kernel/print-fatal-signals
somewhere early in the scripts